Brandon Seah
Brandon Seah
Build database automatically with LTP sequences (https://www.arb-silva.de/projects/living-tree/) instead of SILVA NR Ref database. Motivation is to have a smaller database with authoritative classifications, more suitable for taxonomic classification, cf: https://www.drive5.com/usearch/manual/faq_tax_db.html...
User suggestion: Add option to only perform read filtering and report mapped reads, and not continue further with taxonomic summary and assembly.
Factor in edge case with custom DB when reference sequences too short to report paired end insert size.
User database may contain sequences that are shorter than 800 bp minlength limit hard-coded to the UniVec filtering step in `phyloFlash_makedb.pl`. See issue #115. Allow user to change this option...
In addition to trusted contigs, optionally supply PacBio or Nanopore long reads for assembly step only. This may help in assembly of closely related strains (?).
Bitflag 0x1 not set with SE reads, so no assumptions can be made about 0x40 or 0x80 Need separate subroutine to handle SE reads taxonomy hashing.
With `no-split` option, get error message that 'axis.ticks.margin' is deprecated. Error message below... ``` $ phyloFlash_heatmap.R -v -d -o test01.png --library-name-from-file --no-split ./*.csv Loading required package: optparse Loading libraries Selected...
I'm filing this report for Malin... Heatmap script breaks when only one taxon present in 'Eukaryota'. `hclust` requires at least 2 objects to cluster. Verbose output copied below... ``` $...
- [X] I hereby agree to dual licence this and any previous contributions under both the _Biopython License Agreement_ **AND** the _BSD 3-Clause License_. - [X] I have read the...
**Snakemake version** v8.3.7 **Describe the bug** Attempt to generate a Dockerfile with `snakemake --containerize` fails when the workflow has a global Conda directive. **Logs** The following error message was produced...